This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. Saccadic eye movements remain accurate throughout life despite the changes that occur in the brain during normal development, injury and aging. Here we examine whether this "saccade adaptation" involves the cerebellum. The output element of the cerebellar cortex, the Purkinje cell, discharges two different action potentials: simple and complex spikes (CSs). It has been suggested that CSs report that a movement is in error and this increased activity modifies the frequency of simple spikes so that downstream they alter the signals that drive the eye muscles in a way that reduces the error. We deceive a normal monkey into thinking that its saccades are in error by jumping the target forward or backward as a saccade is launched toward it. The initial saccade undershoots or overshoots the target, respectively, and a second saccade occurs ~150 ms later to bring the eyes on target. During the time that an error exists between the initial and corrective saccade, CSs report the direction and magnitude of the error. If the deception is continued for hundreds of trials, saccades gradually increase or decrease their amplitudes, respectively, so saccades again are accurate. Does the change in CS activity actually drive this adaptation? To address this question, we stimulated in the superior colliculus, the source of the CS activity, to drive the putative adaptation signal artificially. When sub-threshold stimuli are delivered just after a saccade with the target extinguished, there are gradual changes in saccade amplitude, which are identical to those produced by the behavioral target jump paradigm. Moreover, changes in the timing and direction of the artificial error signal produce adaptations like those caused by similar manipulations in target jump paradigm. We suggest that CSs might drive not only saccade adaptation but other kinds of precisionmotor learning as well.